Biomechanical study of failed internal fixation for proximal humerus fractures caused by varus deformity of humeral head
Lu Bai,Honglei Zhang,Peng Chen,Wei Li,Changqing Jiang,Wentao Zhang
DOI: https://doi.org/10.3877/cma.j.issn.2095-5790.2015.02.002
2015-01-01
Abstract:Background Proximal humerus fracture (PHF ) is a common disease in the traumatic orthopedics due to decreased density of bone in the proximal humerus with age .It occurs more often in the medium‐and old‐age people ,accounting for 5% of fractures .With the process of aging population ,the incidence of PHF tends to increase .In recent 20 years ,the surgery methods of unstable PHF change greatly from the percutaneous fixation and common plate for fractures to locking plate fixation with "internal fixator" as mainstay .As the locking plate is widely used for PHF ,some medium‐and long‐term clinical follow‐up data indicated loss of reduction of fracture ,varus deformity of humeral head ,and secondary internal fixation failure were found in PHF cases with two or three sites of defective medial cortex .These complications are characterized by defective medial cortex and varus deformity of humeral head .Over time ,it causes varus and prolapsed of humeral head ,breakage of screws or plate .From biomechanics ,the loss of support for proximal humerus by the medial cortex is an unstable factor .But not all cases with defective medial cortex develop internal fixation failure .The types of internal fixation failure vary in patients .Is the varus deformity of humerus head another risk factor for internal fixation failure? This study investigated the relationship between the varus deformity of humeral head and the internal fixation failures after locking plate treatment for PHF from the biomechanics .Methods Sample treatment and establishment of fracture model:6 pairs of adult humeral samples with antiseptic treatment were obtained from the anatomy department of the Shenzhen University‐affiliated Medical School .The pathological symptoms such as bone disease and chronic fracture were excluded by radiograph .The samples were stored at -20 ℃ before biomechanical study and thawed at room temperature (18 ℃) for 8 h when the biomechanical study was about to start .During thawing ,samples were sprayed with normal saline to prevent drying .Either right or left side of the samples was randomly assigned to the varus deformity group and normal alignment group to establish the fracture model .The fracture model was established on the X‐Y workbench .First ,20 cm was removed from the proximal ends of all samples .The osteotomy was performed in parallel with the vertex tangent of the humeral head at the surgical neck on the junction between humeral metaphysis and the humeral head using oscillating saw .The cortex of the lateral greater tuberosity in proximal humerus was reserved at an amount accounting for 1/3 of the perimeter ,which helped to insert the PHILOS plate .The fracture gap between the osteotomy surfaces was 10 mm .The varus deformity of humeral head was prepared using the method described by Voigt et al .The varus of humeral head by 20° was measured using gauge on the X‐Y workbench .All fractures were fixed using 5‐hole Synthesis PHILOS locking plate .According to recommendation by the AO Organization ,the plate was placed on the greater tuberosity 5‐8 mm away from the intertubercular sulcus ,8‐10 mm away from the greater tuberosity tip .The PHILOS plate was fixed in holes A ,B ,C ,D and E .Three locking screws were used to fix the distal end .All screws were 3 .5 mm locking screws .Procedures of mechanical study :Fixation and testing of samples :The mechanics study was performed on the Instron 8874 (2 kN ,± 1% ) mechanical tester .The fracture model was fixed at the workbench of the tester .The axis of the humeral shaft was vertical to the pressure head of the mechanical tester .The pre‐moulding for mechanical pressure head in the axial compression was prepared using bone cement ,which was matched for the shape of humeral head to avoid damage of the surface of humeral head at the area with concentrating stress during compression .The circular axial compression experiment was performed under the following conditions :minimum pressure of 50 N , maximum pressure of 300 N ,permitted maximum displacement of 5 mm ,cycle number of 5 ,000 .The biomechanical curve was expressed by displacement‐load curve .Determination of internal fixation failure:In the axial compression experiment ,the sample was regarded as internal fixation failure if the variation of the data curve of each sample was within the range of the curve for flexible variables .In the damage mechanical experiment ,real‐time displacement of mechanical tester‐load curve was observed .In the damage mechanical experiment ,the presence of inflection point on the displacement‐load curve when the displacement curve tended to be zero indicated failure of internal fixation .The mechanical measurement was stopped after failure of internal fixation .The samples were removed for observation .The types of internal fixation were recorded .Statistical analysis :The statistical analysis was performed using PASW18 .0 (SPSSInc IBM Chicago ,USA) .The measurement data was expressed as ± s .The difference of maximum load to damage internal fixation in different groups was compared using student t test .The differences in the maximum and minimum displacement of circular mechanics between the varus group and alignment normal group were compared using Mann Whitney‐Wilcoxon . The independent‐sample test was performed for comparison of differences in maximum damage load and maximum displacement of the samples with internal fixation failure between groups .P < 0 .05 indicatedstatisticalsignificance.Results (1)Resultsofcircularmechanics:In5,000axial compressions ,6 pairs of samples did not show internal fixation failure .In measurement of the pressure during axial cycles ,the varus group had greater minimum and maximum displacements than the normal alignment group within 2 ,000 cycles of axial compression ,although no significant difference was found ( P >0 .05) .The mean maximum displacement in the varus group was 2 .37 mm (1 .01‐2 .78) at 3 ,000 cycles of axial compression ,significantly different from 1 .73 mm (0 .91‐2 .08) in the normal alignment group ( P = 0 .044) .After 4 ,000 cycles of axial compression ,the varus group had significantly higher minimum displacement ( P = 0 .039) and maximum displacement ( P = 0 .037) than the normal alignment group .After 5 ,000 cycles of axial compression ,the varus group had a mean minimum displacement of 0 .91 mm (0 .39‐1 .23) and a mean maximum displacement of 3 .08 mm (1 .78‐3 .67) .The normal alignment group had a mean minimum displacement of 0 .56 mm (0 .37‐0 .86) and a mean maximum displacement of 2 .32 mm (1 .58‐2 .94) .The normal alignment group had significantly lower mean minimum displacement ( P =0 .026) and mean maximum displacement ( P=0 .019) than the normal alignment group .(2) Results of damage mechanics:The damage mechanical experiment was performed immediately after the circular load experiment .The mean damage loads were 870 ± 104 N for the normal alignment group and 632 ± 111 N for the varus deformity group . There was significant difference between two groups ( t =3 .812 ,P =0 .003) .In cases with internal fixation failure ,the mean fracture gap displacements were 6 .72 ± 0 .71 mm for the normal alignment group and 5 .67 ± 0 .53 mm for the varus group .There was significant difference between the two groups ( t =2 .994 ,P =0 .013) .Conclusion The varus deformity of the humeral head significantly lowered the stabilization performance of the locking plate internal fixation system for the PHF .